Visit additional Tabor Communication Publications
June 13, 2008
Every June, many of the world's leading experts in high-performance computing convene in Germany to discuss the latest advances in supercomputing and to get a glimpse of what the future holds. Now in its 23rd year, the International Supercomputing Conference, or ISC, was the first international meeting to focus solely on supercomputing.
The conference was created -- and is still led -- by Prof. Hans Werner Meuer of the University of Mannheim, where he was director of the computer center for more than 25 years. Currently, Meuer is a member of the governing boards of several German technology firms and he is still a professor of computer science in Mannheim. He first became involved with supercomputers and parallel computing more than 20 years ago. In 1986, he co-founded and organized the first Mannheim Super Computer Seminar, which has evolved into ISC.
HPCwire got an opportunity to ask Prof. Meuer about the upcoming conference and his thoughts on the direction of supercomputing.
HPCwire: There must be a dozen or more high performance computing conferences during the year now. What do you think distinguishes ISC from the others?
Hans Werner Meuer: We are the HPC conference with the longest tradition, starting already in 1986, i.e., even two years before SCxy in the U.S. Consistently, more than 60 percent and up to 70 percent of our participants attended the previous year's ISC, and the result is that there is an air of familiarity among the participants. It is this feeling of community that adds a personal atmosphere to our gatherings. And I think we try harder to support this atmosphere in the way that we are focusing on making improvements from year to year. This is not only in terms of the professional conference program, but also of the organizational aspects, including the meals and social events. This has been an important consideration from the very beginning in 1986, when we had just 81 attendees.
Since then, ISC has evolved into a high-powered international conference and exhibition that gives a global perspective on the cutting edge of HPC. The conference program tackles hot HPC topics and focuses on future trends and developments. ISC provides the perfect opportunity to see and hear about new technologies. The ISC exhibition with the leading global supercomputing vendors and institutions is a one-of-a-kind platform to see the latest trends and technologies and to meet customers and suppliers. And one very important point, going back to the familiar atmosphere: attendees have the opportunity to talk one-on-one with internationally known experts in the fields of HPC and computational science.
HPCwire: The conference attendance has nearly tripled over the past few years. What do you attribute that to?
Meuer: Obviously, it was always the right time when we moved from Mannheim to Heidelberg in 2001, from Heidelberg to Dresden in 2006 and hopefully also in 2009 when we move to Hamburg.
With each move, we could expand the floor space significantly and therefore could host many new exhibitors and old exhibitors and provide them with a bigger booth space. One of the fastest growing elements of ISC is the exhibition day pass, with which a visitor can take in the full range of exhibits. The attendance for the ISC professional program is also growing. Of course, I like to say this is because of the very compelling set of speakers. But it is also a result of additional pre-conference sessions that will attract more and more participants.
HPCwire: Do you think it could ever get too big for the kind of collegial atmosphere for which ISC is well known?
Meuer: This is a really good question and is one we have asked ourselves many times and discussed at length. For now, and probably for the three years to come in Hamburg, I do not see this really affecting people's decision to attend ISC or not. Of course, you cannot either increase or decrease interest in such a conference as you wish by turning a switch. Please believe me that we will do everything to preserve the collegial atmosphere at ISC events and still maintain the quality of the conference.
HPCwire: What do you think will be highlights of this year's conference?
Meuer: For me, the highlights of ISC'08 include:
The keynote addresses. The opening keynote address by Satoshi Matsuoka is entitled: "Everybody Supercomputes in the Next Generation Cyber-Science Infrastructure." The Thursday keynote presentation by Justin Rattner, Intel Chief Technology Officer, will talk about the second transformation of the computing world in "Multicore/Manycore Platforms Bring Supercomputing to the Masses." And Friday's keynote, which will feature John Salmon of D.E. Shaw Research New York, is about "Toward Millisecond-Scale Molecular Dynamics Simulation of Proteins."
On Friday, we have a panel discussion focusing on a very current topic: "Is HPC Going Green?", especially on how power requirements may be limiting HPC performance.
One of my personal favorites is the unveiling of the latest list of the world's TOP500 supercomputers. The big question this year: will the list contain the first system to break the 1 petaflop/s barrier? [Editor's note: This interview took place right before IBM's announcement that Roadrunner had broken the petaflops barrier.]
One of our most popular features is the review of HPC trends of the past year -- and a look at what's in store for the year to come by noted expert Thomas Sterling.
And one highlight that is unique to ISC consists of two lively sessions during which representatives from the leading HPC vendors find themselves on the "hot seat" as they answer challenging and probing questions. And they have only a short amount of time -- strictly enforced, I might add -- in which to make their points.
Two other highly up-to-date sessions will look at "The European HPC Projects PRACE, PROSPECT and TALOS" and "HPC & Next-Generation Climate Modeling."
HPCwire: Would you ever consider moving the conference to other venues outside of Germany to reflect the international focus of ISC?
Meuer: Yes, we have discussed this already but have not yet come to a final decision. There are pros and cons when moving outside of Germany, of course. The next three years, until 2011, we will definitely stay in Hamburg. After that, it is possible that we would move to another European country.
HPCwire: You mentioned your personal interest in the TOP500 list -- everyone is talking about a petaflops system grabbing the number one spot. What is the significance of this milestone? How does it compare to the teraflops milestone achieved back in 1997?
Meuer: The significance of this milestone is that when reaching this performance level with a real system there is no longer any psychological doubt that you can improve even further. The situation is quite similar to 11 years ago when Intel ASCI Red became the number 1 in our 9th TOP500 list in June 1997 as the first teraflops system on earth. Eight years later in June 2005 -- and predicted by us already in 1997! -- there are only teraflops systems in the TOP500. And now three years after that, in 2008, we definitely will see the first petaflops system in the world. For me, you can compare the breaking of the petaflops barrier with a runner breaking the magical 9.5 second barrier in the 100 meter race.
HPCwire: At a time when the low and middle end of the HPC market is where most of the action is, how do you convince people that the top systems are still relevant?
Meuer: You are right, supercomputing is no longer elitist. Even small and medium-sized companies can benefit from the cluster technology of supercomputers and they are already using this more and more in remarkable ways. But there are many existing problems in science and research, and also in industrial and even atomic weapons applications where you already need petascale systems. In the final talk of the ISC'08 program on Friday, June 20, you can learn from Professor Peter M. Kogge from Notre Dame University about exaflops applications. So, just as we reach the petaflops level, the experts are already looking to exaflops, which equals 1,000 petaflops.
HPCwire: What will the mainstream HPC architecture be in the post-cluster era? Will it be a heterogeneous system architecture like Roadrunner; a low-power, highly-scaled system, like Blue Gene; or something else?
Meuer: I believe that hybrid architectures like Roadrunner are becoming more and more important. By applying Bell's Law to the TOP500 architecture classes, Erich Strohmaier and I have especially looked at "power-efficient systems" classes like IBM's BG series. The next TOP500 lists will show whether or not the "power-efficient systems" class will really be a class of its own. [More about the application of Bell's Law to the TOP500 architecture classes can be found at http://www.top500.org/files/TOP500_Looking_back_HWM.pdf.]
HPCwire: Care to make a prediction when we'll see the first exaflops system on the TOP500?
Meuer: Sure -- I am used to going out on a limb in making such predictions. I predict that we will see the very first Linpack exaflops system in the year 2019. By 2016, we will have only petaflops systems on the list due to our proven performance predictions.
May 16, 2013 |
When it comes to cloud, long distances mean unacceptably high latencies. Researchers from the University of Bonn in Germany examined those latency issues of doing CFD modeling in the cloud by utilizing a common CFD and its utilization in HPC instance types including both CPU and GPU cores of Amazon EC2.
May 15, 2013 |
Supercomputers at the Department of Energy’s National Energy Research Scientific Computing Center (NERSC) have worked on important computational problems such as collapse of the atomic state, the optimization of chemical catalysts, and now modeling popping bubbles.
May 10, 2013 |
Program provides cash awards up to $10,000 for the best open-source end-user applications deployed on 100G network.
May 09, 2013 |
The Japanese government has revealed its plans to best its previous K Computer efforts with what they hope will be the first exascale system...
May 08, 2013 |
For engineers looking to leverage high-performance computing, the accessibility of a cloud-based approach is a powerful draw, but there are costs that may not be readily apparent.
05/10/2013 | Cleversafe, Cray, DDN, NetApp, & Panasas | From Wall Street to Hollywood, drug discovery to homeland security, companies and organizations of all sizes and stripes are coming face to face with the challenges – and opportunities – afforded by Big Data. Before anyone can utilize these extraordinary data repositories, however, they must first harness and manage their data stores, and do so utilizing technologies that underscore affordability, security, and scalability.
04/15/2013 | Bull | “50% of HPC users say their largest jobs scale to 120 cores or less.” How about yours? Are your codes ready to take advantage of today’s and tomorrow’s ultra-parallel HPC systems? Download this White Paper by Analysts Intersect360 Research to see what Bull and Intel’s Center for Excellence in Parallel Programming can do for your codes.
In this demonstration of SGI DMF ZeroWatt disk solution, Dr. Eng Lim Goh, SGI CTO, discusses a function of SGI DMF software to reduce costs and power consumption in an exascale (Big Data) storage datacenter.
The Cray CS300-AC cluster supercomputer offers energy efficient, air-cooled design based on modular, industry-standard platforms featuring the latest processor and network technologies and a wide range of datacenter cooling requirements.